Dafydd Llewellyn

Here's another case where people need to read the relevant design standard. DON'T GUESS. § 23.1093 Induction system icing protection. (a) Reciprocating engines. Each reciprocating engine air induction system must have means to prevent and eliminate icing. Unless this is done by other means, it must be shown that, in air free of visible moisture at a temperature of 30 °F.— (1) Each airplane with sea level engines using conventional venturi carburetors has a preheater that can provide a heat rise of 90 °F. with the engines at 75 percent of maximum continuous power; (2) Each airplane with altitude engines using conventional venturi carburetors has a preheater that can provide a heat rise of 120 °F. with the engines at 75 percent of maximum continuous power; (3) Each airplane with altitude engines using fuel metering device tending to prevent icing has a preheater that, with the engines at 60 percent of maximum continuous power, can provide a heat rise of— (i) 100 °F.; or (ii) 40 °F., if a fluid deicing system meeting the requirements of §§23.1095 through 23.1099 is installed; (4) Each airplane with sea level engine(s) using fuel metering device tending to prevent icing has a sheltered alternate source of air with a preheat of not less than 60 °F with the engines at 75 percent of maximum continuous power; (5) Each airplane with sea level or altitude engine(s) using fuel injection systems having metering components on which impact ice may accumulate has a preheater capable of providing a heat rise of 75 °F when the engine is operating at 75 percent of its maximum continuous power; and (6) Each airplane with sea level or altitude engine(s) using fuel injection systems not having fuel metering components projecting into the airstream on which ice may form, and introducing fuel into the air induction system downstream of any components or other obstruction on which ice produced by fuel evaporation may form, has a sheltered alternate source of air with a preheat of not less than 60 °F with the engines at 75 percent of its maximum continuous power. (b) Turbine engines. (1) Each turbine engine and its air inlet system must operate throughout the flight power range of the engine (including idling), without the accumulation of ice on engine or inlet system components that would adversely affect engine operation or cause a serious loss of power or thrust— (i) Under the icing conditions specified in appendix C of part 25 of this chapter; and (ii) In snow, both falling and blowing, within the limitations established for the airplane for such operation. (2) Each turbine engine must idle for 30 minutes on the ground, with the air bleed available for engine icing protection at its critical condition, without adverse effect, in an atmosphere that is at a temperature between 15° and 30 °F (between −9° and −1 °C) and has a liquid water content not less than 0.3 grams per cubic meter in the form of drops having a mean effective diameter not less than 20 microns, followed by momentary operation at takeoff power or thrust. During the 30 minutes of idle operation, the engine may be run up periodically to a moderate power or thrust setting in a manner acceptable to the Administrator. © Reciprocating engines with Superchargers. For airplanes with reciprocating engines having superchargers to pressurize the air before it enters the fuel metering device, the heat rise in the air caused by that supercharging at any altitude may be utilized in determining compliance with paragraph (a) of this section if the heat rise utilized is that which will be available, automatically, for the applicable altitudes and operating condition because of supercharging. [Amdt. 23-7, 34 FR 13095, Aug. 13, 1969, as amended by Amdt. 23–15, 39 FR 35460, Oct. 1, 1974; Amdt. 23–17, 41 FR 55465, Dec. 20, 1976; Amdt. 23–18, 42 FR 15041, Mar. 17, 1977; Amdt. 23–29, 49 FR 6847, Feb. 23, 1984; Amdt. 23–43, 58 FR 18973, Apr. 9, 1993; Amdt. 23–51, 61 FR 5137, Feb. 9, 1996]

OK, leave the questions of which control surface and which aircraft out of it. Look at it as a lesson for design. Control surfaces CAN become disconnected, due to a control system failure. It happens; I've had one myself, and was directly involved in rectifying the cause of another. So, what can one do to prevent a disconnected control surface from fluttering? Answer: Mass-balance it correctly* . And design it to "float" in a neutral position, when disconnected. Unfortunately, the very tight weight limits of recreational aircraft mean that very few of them can afford the weight involved in correctly mass-balancing; to be effective, the balance mass needs to be at quite a short arm, so you end up increasing the total weight of the control surfaces by around 300%. So reliance is placed instead on control system stiffness and freedom from slack, and in the case of push-pull cable, the damping provided by the cable friction, in conjunction with very light control surfaces. In these aircraft, the only answer is to be very vigilant about control system integrity. Also, Frise-type ailerons cannot be aerodynamically balanced singly; they work as a pair, and therefore are quite likely to go hard-over if disconnected. Elevators are often equipped with fixed tabs that drive the elevators upwards, as a means of improving the stick-free stability; that will result in them free-floating in a position suitable for flight at close to the stall speed - so the elevator really does need to be driven by the trim system quite independently of the main pitch control system. * Correct mass balance practices generally require that the balance mass be biased towards the tip of the wing, tailplane or fin. Also, there are stringent requirements on the strength and stiffness of the attachment of the balance mass. This is an area where the design standards definitely need to be consulted.

You know, you could actually do this yourselves. It would amount to a peer group assessment. You could design the list of questions to be addressed, and the panel could be any three members of RAA who had no interest in or connection with the builder, the supplier, or the aircraft. Self-help, in fact. Or is this too radical a concept for a bunch of whingers?

This aspect overlooks the fact that CASA DOES have a responsibility, in regard to experimental aircraft, to protect other airspace users and people on the ground. If you people would do as I ask and READ THE REFERENCE - in this case, CASA AC 21.10 - you will see how this is done. If Joe Bloggs turns up with an experimental aircraft, all the instructor has to do is look at the limitations on the experimental certificate; those limitations are there to protect other airspace users and persons on the ground. So having ascertained that the proposed flight will abide by those limitations, what remains is for the instructor to assess the risk to himself. That's where a risk analysis score would come into play. What I do NOT know is whether RAA applies the same procedures as are given in AC 21.10, to -19 registered aircraft. It would surely be negligent if it does not. For the benefit of readers who have not started far enough back in this thread, the suggestion was that RAA appoints an expert panel, to perform a risk analysis, that was in effect an extension of the risk analysis concept that appears in the appendices of AC 21.10. It would contain a list of questions, with a score for each answer. The total score would indicate the risk assessment for the aircraft. As a builder, there would be no compulsion to have such an analysis run, but it would be in your interests, whether as a builder or a kit supplier, to get a high score. So it's a voluntary thing, but it's done by a third party, and a high score would result in reduced limitations on the use of the aircraft.

People need to (preferably) do their flight checks on the aircraft they normally fly. It's up to the instructor. You could approach the issue of giving the instructor a basis on which to make his decision, by setting a minimum score in the risk analysis.

I agree it's something that everybody in recreational aviation needs to understand. However it's a common mistake that aircraft design standards are mainly about structures; that's only one facet of it - and I suspect the aspect that causes the least accidents. I did say that people tend to state what they think the design standard requires, instead of reading it to find out what it actually calls for; that's a typical example. The problem I see with most of the designs of recreational aircraft, is the fundamental philosophy; this is easiest to illustrate by the fact that the structures are almost always "single load path" designs - and so they cannot withstand a single failure anywhere. This is equally true of the control systems, the undercarriages, etcetera. The service bulletin on bulkhead nuts on push-pull cable systems illustrates this. Light aircraft design standards require redundancy (i.e. the ability to withstand a single failure) only in regard to having dual ignition on the engine, and a back-up electric fuel pump in the fuel system, and (in the case of FAR 23) the ability to land the aircraft using the elevator trim system in lieu of the primary elevator control. Lately, one finds some mention of the ability to make gentle turns using only the rudder. This practice arose, in the case of structures, because it was very complex to analyse a redundant structure, before computers came into general use; but nowadays, with Finite Element Analysis, it's fairly straightforward to do so. A redundant structure may weigh only a few percent more than a single load-path one, but the extremely tight weight limits for recreational aircraft largely preclude that extra weight. Airliners invariably have redundant control systems - one of the advantages of conventional cable systems is that a broken cable is not prone to jamming in the structure, as a pushrod is. So you find duplicated control runs to all the flight controls - the upper fuselage of a Fokker Friendship looks like the inside of a piano, with all the cables. That extent is not necessary in a recreational aircraft - but we need, I believe, to start thinking about failure modes for structures and control systems, etc, and learn how to do "smart" design to cope with them or eliminate them.

Tailwheel aircraft generally have better performance than their equivalent nosewheel bretheren; and vastly better propeller clearance when taxiing; and - believe it or not - better crosswind capability. They definitely have their place. As a CAR 35, I've seen lots of damaged fuselages due to nosewheel loads. The history of the early Jabiru nosewheel is informative; it started out meeting FAR 23 drop-test criteria - and people kept breaking it. It was beefed-up in several increments, and ended up twice as strong as FAR 23 requires; Rod Stiff commented that every increment in strength seemed to bring out another level of pilot ham-fistedness. It's now at a strength such than the fuselage breaks instead. I've also been asked for opinions in regard to prop strike (See CASA AD/ENG/6 Amdt 1) for scores of nosewheel aircraft - but almost never for tailwheel aircraft. Yes, they require a little higher level of pilot skill - so not the "flying Chevrolet" epitome. However, I definitely prefer them - and so do most "bush" pilots.

Yes, that's the reality of it. I don't have an issue with people doing that; however I think the pathway to improved understanding can be made easier.

Thanks, BR; now to find one compatible with mineral hydraulic fluid . . .

In point of fact, liability issues under the Trade Practices Act and Negligence in Tort both act to make it virtually impossible to impose rules or guidelines on experimental aircraft. As soon as any authority imposes ANY rule, Pandora's box opens and the lawyers step in. This was hammered out flat in the process of drafting CASR Part 21; as Middo will remember, you cannot have "a little bit of airworthiness" in any formal sense - either it is airworthy (i.e. it complies with the relevant standard fully), or it ain't. An experimental certificate is in reality a certificate of non-airworthiness - which is a legal concept - tho it does not mean the thing is not fit to fly - which is a physical state of the aircraft. However, a risk analysis approach functions to assist people to understand the nature of the risk they are accepting, and it's about the only way I can see to do that without opening the box.

Andy, there's a quite strong "We don' need no steenking engineers!" sentiment out there. In the aviation mainstream, professional engineers go over the design against the design standard, line by line, and write reports to show the thing complies. CASA reads the reports, and there is often a ding-dong argument - but the end result can only be that CASA is truly satisfied that the aircraft complies with the design standard. That process takes several man-years and costs a lot of money. The philosophy behind DIY experimental aircraft (however described) is diametrically opposite to that; it allows people to slap anything together and fly it - the only constraint being where and under what conditions they may do so. So the whole thing is opposite. Judging by the responses I keep seeing on this website, almost nobody in rec. aviation has ever even glanced at one of the design standards. They all state what they imagine the requirement should be, without ever actually looking it up. What I am advocating is using the established standard as guidance material for a risk analysis. The sort of question/answer that could be developed might include such matters of fact as "Does the elevator trim system drive the elevators independently of the main elevator control system (yes/no)" - etcetera.

May I suggest that a way to approach this, for a scratch, plans, or "major content" kit, is for the builder of such an aircraft to acquire a copy of FAA AC 43.13-1 BEFORE HE STARTS, and also he should download a relevant design standard - they are all fairly similar - such as FAR Part 23, and read the relevant section of it as he builds each part, especially the sections on "Design and Construction" and "Engines". He doesn't need to be able to use the formulae on the structures part of it; just read the parts that are in plain english - which is most of it, actually. The devil is almost always in the detail. These references will go a long way to showing what details are important. It would be good if the "risk analysis" mechanism can be set up so the kit supplier does this work, but in the meantime, anybody can do the above.

Wrong: § 23.735 Brakes. (a) Brakes must be provided. The landing brake kinetic energy capacity rating of each main wheel brake assembly must not be less than the kinetic energy absorption requirements determined under either of the following methods: (1) The brake kinetic energy absorption requirements must be based on a conservative rational analysis of the sequence of events expected during landing at the design landing weight. (2) Instead of a rational analysis, the kinetic energy absorption requirements for each main wheel brake assembly may be derived from the following formula: KE=0.0443 WV 2 /N where— KE=Kinetic energy per wheel (ft.-lb.); W=Design landing weight (lb.); V=Airplane speed in knots. V must be not less than VS√, the poweroff stalling speed of the airplane at sea level, at the design landing weight, and in the landing configuration; and N=Number of main wheels with brakes. (b) Brakes must be able to prevent the wheels from rolling on a paved runway with takeoff power on the critical engine, but need not prevent movement of the airplane with wheels locked.

I've always found that the most effective short-field technique with a taildragger is to dump the flaps at (or an instant before) touchdown, in a full 3-point attitude. That puts it firmly on the ground, and you can then use the brakes quite heavily if you need to. The same technique also works in tricycle undercarriage aircraft, provided you can hold the nosewheel off and put it down gently. For that reason, I dislike electric flaps, in a tailwheel aircraft; the inability to dump the flaps cripples their utility. However, this technique does require that you be able to judge your hold-off height accurately, and that you know when it is going to stop flying; so it's not taught to ab-initio students - or not taught at all. Also, you need to be very careful doing this in an aircraft with retractable undercarriage; it's far too easy to dump the undercarriage instead of the flaps. QUOTE: Would probably like to see something more "proportional", say from tail wheel on the deck to horizontal rather than an on/off situation. The idea sounds nice, but it's very difficult to achieve with any sort of stone-axe simple system. Also, I don't think it would work if the effect it delayed until the aircraft is horizontal; by that stage, it's almost on the point of falling over, and it will have considerable overturning momentum. What is needed, I suggest, is something that acts much faster than any pilot can react. A small effect at the right time, rather than a large effect, almost too late.

Not if you want to get a Type Certificate. Take a look at what is involved in certificating software. In a certification exercise, the KISS principle applies.

Try a couple of aspirin . . .

Hmmm. I once had to land a Jabiru with a disconnected elevator, because a cable outer anchor let go. In the Jab, you can land it using the elevator trim. It landed without a scratch. I think there's more than one message, there. FAR 23 requires that the aircraft can be landed with no elevator input - which is not quite the same as a disconnect, but getting close. That's the sort of thing I mean about the importance of understanding the design standards.

Well , I was not talking about ABS, tho the thread drifted off onto it. I'm talking about a simple pressure-relieving anti-locking valve of the simplest type, that drops the brake mechanical advantage by (say) 50% when the tail wheel comes off the ground.

If you're building from plans or a kit, AND the original aircraft from which the kit was derived, had a Type Certificate, then at least there's a background that the original design was thoroughly vetted by an appropriate third party, to verify that it complied with the relevant design standard. The design standards for recreational aircraft are not completely beyond criticism, but they are a big improvement on nothing. Bear in mind that aircraft design standards are always MINIMUM standards; they don't guarantee excellence, but they are intended to provide a tolerably safe product provided you use it within its limits, and they ensure those limits are specified. However, a kit under the "major content" rules can be supplied by anybody; and it does not have to have been built and tested. There's no formal control of it whatsoever. Not of its design, nor of the materials supplied in the kit. You're right out on your own. So if you tackle one of these, you really need to have a lot of aviation knowledge, and every bit of help you can get. It can be done but the potential risk is high, unless you really know what you are doing. In practice, the range can lie anywhere between these extremes. An LSA kit is a half-way house; the original aircraft was self-certified by its manufacturer, and before he can do that, he has to either hold a Production Certificate (which means he has previously manufactured a type certificated aircraft, and has the necessary production quality control system in place) - OR he has to employ or otherwise use recognised aviation professionals with the appropriate qualifications. However, that is enforced after the event by the normal criminal processes that apply to fraud, so it can be abused (for a while). I covered all this under the thread "Caveat Emptor".

Thanks to you all for the input. All the points you raise are valid; however passive rear wheel anti-locking systems have been around in the automotive scene for at least forty years now, so I think the reliability aspect isn't really a problem. First cost isn't likely to be much of an issue either, if one can use an off-the-shelf automotive part, and there are a large variety to choose from. You all seem to accept that it could be done without any significant adverse effect on the ground handling - and that's the point that concerned me. I've no current commercial interest in the idea - feel free to use it, so far as I am concerned. I'd put it on my own tail dragger.

Yes, I agree fully.

In general, you are right; however I could quite easily lift the tail on my Auster if I used the brakes while they are cold, when taxiing slowly - and I had to be careful using power to taxi over a bump at the edge of the tarmac, too. As soon as the brakes warmed up, they lost the power to do that. American aircraft generally have their main wheels set a little further forward than English aircraft of that era; so much so, in the case of the Pawnee, that it would "hobby-horse" if you were not careful - which tended to break the shock-absorber, if it happened too often. However, that means that on the rare occasion when they do start to lift the tail, things happen very fast.

It's always worked for me, especially if you turn into wind before running up; however, I saw a Indonesian air-force Colonel stand the Seabird Seeker on its nose because he had never previously operated a tail-wheel aircraft; and I've seen a very experienced tailwheel pilot put a Bellanca Scout right over on its back, because a dust-devil formed just behind the aircraft during run-up, which negated the effect of holding back stick. It doesn't only occur at run-up, either; the Cessna 170 or whatever that flipped in the video may well not have done so had it had a device of this nature. I'm looking at something that can be added, to reduce the insurance hull rate for tail wheel aircraft. Simply teaching people to use correct run-up technique is of course essential; but it won't alter the actuarial facts that the bean-counters rely on. It needs something new to do that.

I've no objection to providing the data either; and how you use it is up to you; my analogy of an expert witness was not intended to suggest that you are aiming to use it in any adversarial manner - merely that the sort of information it seems to me that you would need is not dissimilar. I would merely observe that any person's point of view depends on the whole of his experience, and merely focussing on (say) pilot licence qualifications is likely to be only a small facet of that - and therefore, misleading.

Where I disagree, I've generally said so. If I don't say so, you can assume I have no particular issue. None of us is right all the time . . .